oci runtime error ,ldcache error : process /sbin/ldconfig failed with error code 1

[codexie]

Hi,
I have installed nvidia-docker2 ,cuda9.0 and nvidia-driver384.81,and the GPU is Tesla M40 . However,when I exec "docker run --runtime=nvidia --rm nvidia/cuda nvidia-smi" , the error occurs as follows :
Error response from daemon: oci runtime error: container_linux.go:265: starting container process caused "process_linux.go:368: container init caused "process_linux.go:351:
running prestart hook 1 caused \"error running hook: exit status 1, stdout: , stderr: exec command: [/usr/bin/nvidia-container-cli --load-kmods configure --ldconfig=@/sbin/ldconfig.real
--device=all --compute --utility --require=cuda>=9.0 --pid=9152 /data/docker/aufs/mnt/a5580b0483a1324f820c340aac1c491eb869899d3e868ec51bd99c2a66a18791] \\nvidia-container-cli:
ldcache error : process /sbin/ldconfig failed with error code 1 \\n\""".

 What can be the problem for the error and how to fix it ?
 Thank You

[flx42]

What's your OS distribution?

[codexie]

Thanks for reply. The os informations: Linux version 3.10.106, (gcc version 4.8.2 20140120 (Red Hat 4.8.2-16) (GCC)) , CentOS Linux release 7.2 (Final) .

[3XX0]

Please follow the issue template when filling an issue, especially the last 4 bullet points

[3XX0]

Are you using a custom kernel/distribution?
Your output is weird, it shows /sbin/ldconfig.real and AUFS which suggest you are on an Ubuntu distribution but also Centos7/RHEL7.

What's the output of:
cat /boot/config-$(uname -r) | grep -i seccomp

[flx42]

ping @codexie

[codexie]

yes ,our company use custom kernel based linux kernel 3.10, it says doesn't support seccomp mode temporarily.Hope the problem is this,and the new kernel will support it soon , then I try it again.

[3XX0]

Any update @codexie ?

[codexie]

Yes, when the kernel updated to support seccomp mode ,it works done.
Thanks a lot all of you
Best Wishes.
@3XX0 @flx42